April 20 through April 26

For Lecture, Monday, April 22

This is the last lecture on file systems; it covers distributed file systems. In preparation, read Tanenbaum chapter 13 through section 13.2.4.

For Recitation, Tuesday, April 23

Read "Disconnected operation in the Coda file system" by Kistler and Satyanarayanan, reading #32. This paper is the second paper in 6.033 that discusses issues related to mobile computing. The first one talked about mobile IP (reading #17); the Coda paper looks at issues when a mobile computer disconnects and reconnects from a file server. Everyone calls Satyanarayanan "Satya."

Your one-page reading report should address the following question:

The paper mentions that scalability was an a priori consideration in the Coda system, as opposed to an afterthought. What mechanisms and policies does Coda implement to achieve scalability, and what are the trade-offs for making Coda largely scalable?

For Lecture, Wednesday, April 24

The first lecture of the last technical topic in 6.033. The topic of this set of lectures is fault-tolerant computing systems. In preparation, read "Chocolate" by Plauger (reading #36) and "Engineering: history and failure" by Petroski (reading #35). These two papers are very short, but provide you with some good insights. Good easy reading. In addition, start reading the paper assigned for Thursday, April 25, reading #33: "High-availability computer systems" by Gray and Siewiorek; in particular, read until the section "Fault-tolerant design concepts" so that you can follow the terminology used in lecture.

For Recitation, Thursday, April 25

The reading for today is "High-availability computer systems" by Gray and Siewiorek, reading #33. This paper is another heavy duty paper; read it with care. In addition, read "The space shuttle primary computer system" by Gifford and Spector, reading #34. This paper is quite long, but you do not have to understand the technical details of the system discussed; we mostly read it to get a sense of what is involved in building a critical computer system. This paper is the first of a series of papers that report on critical computer systems. In the next couple of weeks we will read two others.

For Special Lecture, Friday, April 26

Jeff Schiller, the current network manager for MIT and one of Athena's architects, will give a lecture on his experiences with distributed systems. Jeff is known to give very exciting talks with many insights.


System aphorism of the week
An engineer is a person who can do for a dime what any fool can do for a dollar. (Anonymous)

6.033 Handout 27, issued 4/19/96